Major inflation events

The inflation of 2021 to 2023

This is the one nearly everybody reading has lived through. All 108 countries here recorded a peak between 2021 and 2023. For 44 of them it was the highest reading their series has ever held, and the median peak across all of them was 10.8%. In October 2022 alone, 15 of them reached their worst month.

The synchronisation is the clue to the cause. Countries with different currencies, different central banks and unconnected economies do not peak within weeks of one another by coincidence; they do it when the shock is shared. Three arrived in succession.

The first was supply. Lockdowns closed factories and ports, and the shipping that connects them fell out of sequence, so goods took longer to make and longer to arrive. The second was demand. Governments had spent heavily to support household incomes through those closures, and with services shut much of that money went unspent; when restrictions lifted it met an economy that could not yet produce enough to satisfy it. The third was war. Russia's invasion of Ukraine in February 2022 disrupted two of the world's largest suppliers of wheat, fertiliser, oil and gas at once, and the price of everything grown, heated or transported rose with them.

Central banks were slow to respond, in most cases describing the rise as transitory well into 2021, and then raised interest rates faster than at any time since the 1980s. Inflation fell about as quickly as it had climbed. The last stretch back to target has proved slower than the descent, which is the usual pattern.

The oil shocks of the 1970s

In October 1973 the Arab members of OPEC cut production and embargoed exports to countries that had supported Israel in that month's war. The price of crude roughly quadrupled. A second shock followed the Iranian revolution in 1979.

The effect depended almost entirely on how much oil a country imported and how its wages were set. Japan, importing very nearly all of its energy, reached 25.0% in the year to February 1974, the highest figure in a series running back to 1970. Britain, where pay settlements chased prices and prices chased pay, reached 26.9% in the year to August 1975 on the RPI measure. The United States peaked later and lower, at 14.6% in the year to March 1980, but stayed elevated for fifteen years.

What the decade established was that an oil price is a tax on everything, and that a country whose wages automatically follow prices turns a one-off shock into a spiral.

Hyperinflation, and what causes it

Hyperinflation has a definition rather than just a feeling. The usual one, set by the economist Phillip Cagan in 1956, is prices rising by more than 50% in a single month. That is a far higher bar than it sounds: sustained for a year it multiplies prices roughly a hundredfold. Most of what gets called hyperinflation in ordinary speech, including Turkey's 125.9% in the year to January 1995, is chronic high inflation, which is a different disease with a different cure.

Two episodes in this data clear the bar. Argentine prices rose 196.6% in the single month to July 1989. Chilean prices rose 87.6% in the month to October 1973.

The cause is almost always the same, and it is fiscal rather than monetary in origin. A government commits to spending it cannot raise in tax, and finds it cannot borrow the difference either, because lenders have stopped or were never there. The remaining option is to have the central bank create the money. This works briefly, because the public accepts the new notes at something close to their old value. It stops working as soon as people realise it will continue, at which point they spend money as soon as they receive it rather than hold something that is losing value by the day. That circulates each note faster, which raises prices further, which means the government must print faster still to fund the same real spending. The spiral is driven by expectation, which is why it can accelerate long after the original deficit has been addressed.

Germany in 1923 is the clearest illustration, and it is not in this data: the currency was abandoned in November of that year, by which point a United States dollar bought 4.2 trillion marks. The war had been financed by borrowing rather than taxation, on the expectation that the defeated would pay for it, and reparations followed defeat instead. When Germany fell behind on reparations, French and Belgian troops occupied the Ruhr on 11 January 1923. Berlin called for passive resistance and undertook to pay some two million striking workers, which meant printing money to fund them while simultaneously losing the tax revenue of the country's industrial heartland. That is the mechanism in its purest form: an obligation the state could not fund, met by the printing press.

Argentina's case was slower and its causes were accumulated rather than sudden: decades of deficits covered by the central bank, an external debt crisis through the 1980s that closed off borrowing, and widespread indexation that passed each shock straight through to wages and contracts. Chile's was different again. Prices had been held down by decree while the government ran very large deficits, so the inflation existed but was not being measured; when the controls came off after the coup of September 1973 the accumulated pressure arrived at once, which is why the worst single month sits immediately after it.

Every one of these ended in the same way. Not with better policy within the existing system, but with a new currency and a rule that removed the government's ability to finance itself by creating money: Germany's Rentenmark in November 1923, Argentina's Convertibility Law in April 1991. The credibility of the rule mattered more than its mechanics.

How countries got inflation down

The answer most countries eventually reached has two parts, and both are institutional rather than economic. The first is to give the central bank a single published number to aim at. New Zealand did it first, in 1990, and most of this site's 108 countries now have a target; each country's page states its own beside its current rate.

The second is to put the bank beyond the reach of the government that would like lower interest rates before an election. Germany built the Bundesbank that way in 1957 and insisted the European Central Bank be built the same way. Chile made its central bank autonomous in October 1989 under a law written to be hard to repeal. Britain handed the Bank of England the power to set rates in May 1997. The Federal Reserve, founded in 1913, did not adopt a numerical target until January 2012.

Neither part guarantees anything, as 2022 showed. What they change is what happens next: an inflation that is expected to be brought back to a stated number behaves differently from one that nobody has promised to stop.